Skip to content
/ pupdb Public

A simple file-based key-value database written in Python

License

Notifications You must be signed in to change notification settings

tuxmonk/pupdb

Repository files navigation

PupDB Logo

Build Status codecov PyPI version fury.io Supported Python Versions

What is it?

PupDB is an ernest attempt to create a simple file-based key-value database written in Python.

Why PupDB?

The objective behind the creation of PupDB is to create a database system which performs simple persistence operations well and data can be accessed with a minimalist, easy-to-use API with least configuration.

PupDB is the best choice when:

  1. You need a simple NoSQL data store with an interface as simple as a Python dict, and want to start storing and retrieving data within a few minutes.
  2. You want to create an application without bothering much about configuration and setup for storing data.
  3. Your database is not very huge i.e. not greater than a few megabytes of data.

When not to use PupDB:

  1. You want to perform advanced queries on your data.
  2. Your database is larger than a few megabytes of data.
  3. You want a database software that supports advanced capabilities like indexing, partitioning etc.

Salient Features

  1. Multi-processing support: Same database file can be used across multiple processes concurrently.
  2. Mult-threading support: Same database file (with separate PupDB instance per thread) can be used concurrently.
  3. REST-based HTTP Interface: Apart from using it as a python package, you can also use PupDB via a flask-based HTTP interface. This way you can use PupDB with programming languages other than Python.

Installation

PupDB can be installed using pip:

pip install pupdb

Basic API Documentation and Usage

PupDB can be instantiated as follows:

from pupdb.core import PupDB

# Specify database file path as an argument to the PupDB constructor. That's it.
db = PupDB('db.json')
  1. set(key, value): Stores the value mapped to key in the database file.
db.set('test_key', 'test_value')
  1. get(key): Returns the value mapped to key in the database file. Returns None if key is not found.
db.get('test_key')
  1. remove(key): Removes the key from the database file. Raises a KeyError if key is not found in the database file.
# Remove the key `test_key` from the db.
db.remove('test_key')

# Try doing the same again and it'll raise a `KeyError`,
# as the key has already been deleted from the db in the above step.
db.remove('test_key')
  1. keys(): Returns the keys present in the database file. Return type is list in Python 2 and Dictionary view object (similar to dict.keys()) in Python 3.
# Python 2
print db.keys() # This will print ['key1', ...]

# Python 3
print(list(db.keys())) # This will print ['key1', ...]
  1. values(): Returns the values of all keys present in the database file. Return type is list for Python 2 and Dictionary view object (similar to dict.values()) in Python 3.
# Python 2
print db.values() # This will print ['value1', ...]

# Python 3
print(list(db.values())) # This will print ['value1', ...]
  1. items(): Returns the values of all keys present in the database file. Return type is list for Python 2 and Dictionary view object (similar to dict.items()) in Python 3.
# Python 2
print db.items() # This will print [('key1', 'value1'), ...]

# Python 3
print(list(db.items())) # This will print [('key1', 'value1'), ...]
  1. dumps(): Returns a json dump of the entire database file sorted by key.
db.dumps() # This will print the database json string.
  1. truncate_db(): Removes all data from the database file i.e. truncates the database file.
db.truncate_db()
print(db) # Will print an empty database dict '{}', as the database has been truncated.

Using the PupDB HTTP/REST Interface

Using the HTTP/REST Interface, all PupDB-related operations can be performed without using PupDB as a Python package. As a result, PupDB can be used in any programming language that can make HTTP requests.

To start PupDB's gunicorn-based flask server:

from pupdb.server import start_http_server

# Start the gunicorn server (with 4 worker threads).
start_http_server()

The server will listen to local port 4000. The server will be available at http://localhost:4000.

HTTP API Endpoints

  1. /get?key=<key-goes-here> (Method: GET): This API endpoint is an interface to PupDB's get() method. e.g.:
curl -XGET http://localhost:4000/get?key=test

The above curl request will fetch the result for key test.

  1. /set (Method: POST): This API endpoint is an interface to PupDB's set() method. e.g.:
curl -XPOST http://localhost:4000/set -H 'Content-Type: application/json' -d '{"key": "test", "value": "1234"}'

The above curl request will set the value 1234 to key test in the database.

  1. /remove/<key-goes-here> (Method: DELETE): This API endpoint is an interface to PupDB's remove() method. e.g.:
curl -XDELETE http://localhost:4000/remove/test

The above curl request will remove the key test in the database. It returns a 404 Not Found if the key does not exist in the database.

  1. /keys (Method: GET): This API endpoint is an interface to PupDB's keys() method. e.g.:
curl -XGET http://localhost:4000/keys

The above curl request will return a payload containing the list of keys in the database.

  1. /values (Method: GET): This API endpoint is an interface to PupDB's values() method. e.g.:
curl -XGET http://localhost:4000/values

The above curl request will return a payload containing the list of values of all keys in the database.

  1. /items (Method: GET): This API endpoint is an interface to PupDB's items() method. e.g.:
curl -XGET http://localhost:4000/items

The above curl request will return a payload containing the list of [key, value] pairs in the database.

  1. /dumps (Method: GET): This API endpoint is an interface to PupDB's dumps() method. e.g.:
curl -XGET http://localhost:4000/dumps

The above curl request will return a payload containing the string dump of the entire database.

  1. /truncate-db (Method: POST): This API endpoint is an interface to PupDB's truncate_db() method. e.g.:
curl -XPOST http://localhost:4000/truncate-db

The above curl request will truncate i.e. remove all key-value pairs from the database.

Versioning

We use SemVer for versioning. For the versions available, see the tags on this repository.

License

This project is licensed under the MIT License - see the LICENSE.txt file for more details.